Near singular matrix

[marg.c.f.]

Hello,

I am doing my thesis and I need some help because I am getting the "Near singular matrix" error once I try to run the model.

I am studying the impact of a fund on its portfolio of companies' performance, using undated panel data. Undated because I am using 1, 2, 3 (...) years after the fund's investment.

My model is: the dependent variable as companies'performance and the independent variables as age, a dummy for receiving the fund, dummies to indicate the companies' size (micro, small and medium), industry dummies, year dummies and location of the company dummies.

I performed the correlation matrix to check if there is any almost perfect correlation between the independent variables and I haven't found any. The maximum value I found was of around 0,4, which I found between the different location dummies and also between year dummies.

Any help would be great since this is the first time I am working with panel data and my thesis supervisor has no knowledge about this subject.

Thank you.

[startz]

Check for the "dummy variable trap."

[marg.c.f.]

Yes, it was the problem. I solved it by excluding one dummy in each category since I basically included all possibilities in a category, thanks!